Physical disability is the most prevalent major health problem of aging populations, and the associated needs for medical services are enormous. This longitudinal study of University of Pennsylvania alumni aims to identify risk factors for physical disability, associated use of medical services, and costs of care; to assess trends and changes in risk factors from young adulthood through old age as they relate to changes in physical function and costs of care; and to examine the consequences of risk factor changes in terms of cumulative disability (compression of morbidity) and health care utilization over time. The objective is to understand how and when improvements in risk factor status may prevent or retard functional decline and reduce health care utilizing and costs. Data were collected on this cohort by medical exam in the 1930's and by follow-up questionnaires in 1962, 1976, and 1980. The current study of disability began in 1986, when 2634 men (77%) and women (23%) average age 68, were recruited. To date 1692 survivors have been surveyed annually. This proposal requests five years of funding to add a new health care cost component and to continue data acquisition into the ages when this cohort will experience the greatest degree of disability, declines in health, and use of health services (the 80's). The study will address issues for which data are currently very limited, e.g., risk factors for disability and health care costs in the oldest- old (80+years), lifetime disability and health care utilization, changes in predictors from young adulthood through old age, and the degree to which improvements in risk factor status are associated with decreased disability and costs. Information will be collected on physical disability, including limitations in ADL and IADL, use of aids and help from others, and direct and indirect health care costs, including hospital and nursing home stays, outpatient and emergency room visits, diagnostic and laboratory tests, non-traditional treatments, home care, medication use, inactivity days, and work loss days. Predictor variables include demographics, income, insurance, weight, smoking, alcohol, physical activity, social activity, social support, comorbidity, fractures, falls joint pain, visual and cognitive impairments, depression, health locus of control, and negative life events. Results will suggest public policy initiatives aimed at decreasing morbidity and medical care costs.